Adolescents suffer many acute adverse consequences from alcohol use. Recent epidemiological findings have linked adolescence to the development of alcohol abuse and dependence. This co-occurrence suggests that the processes of adolescent development may encourage the development of alcohol involvement. This project investigates this possibility during entry into adolescence because: 1) this age range may be a developmental "hot zone," in which a number of biological, psychological, and environmental influences on drinking converge;2) we believe our plan to frequently assess alcohol use will open a new "window" on drinking development without incurring cost and feasibility issues associated with an earlier start and lengthier investigation;and, most importantly, 3) our previous research on alcohol expectancies directs us to this period as central to understanding drinking onset. In studies of expectancies in children before drinking begins, we observed a cognitive shift with the transition to adolescence: before adolescence, children primarily associated negative outcomes with drinking, but by early adolescence, drinking was primarily associated with positive and arousing outcomes. Our intent is to identify the source of this shift as a means of probing key developmental influences on drinking. Using an accelerated longitudinal/sibling design, we will recruit 245 families, each with a child 8.5 years of age and a same-gender older sibling whose age is within three years of the index participant. Both parents and both children will be assessed three times per year for 4-5 years, to span 7 developmental years. Our aims are to use growth modeling to: 1) measure selected biopsychosocial developmental changes as participants move into adolescence;2) determine which of these changes best predicts the change in expectancies from primarily negative to primarily positive;3) determine which of these changes best predicts the development of drinking;and 4) test which parameters of the adolescent transition influence the development of drinking via mediation by expectancy change. Identification of such pathways should clarify which developmental processes are not just correlated with drinking, but actually contribute to drinking involvement through the expectancy mechanism. Increased understanding of these causal pathways will direct us to the most important targets for intervention, and may suggest new mechanisms that may be modified to delay/prevent drinking onset.